Yuval Noah Harari recently sat down with Ezra Klein for two hours. They talked about liberalism, AI, nationalism, war, and the technology that is reshaping all of it.
It is a thoughtful conversation. It is also, from beginning to end, a description of a supply problem that neither man names.
Harari diagnoses. Klein nods. The mechanism is missing.
Here is Harari on social media:
The algorithms experimented on millions, on billions, of human guinea pigs to see how we make humans more engaged. They discovered that the easiest way to make people engaged is to press the hate button or the greed button or the fear button.
That is not a description of a discourse problem. It is a description of a product. Built by a supplier. Tested on buyers. Sold without disclosure. The cost of that product, measured in teen mental health, in election integrity, in social trust, did not appear on any receipt.
It appeared on the bill.
Harari knows this. He says it plainly. What he does not say is who should be writing the terms of the next product before the same thing happens again.
He gets closer when the conversation turns to AI.
The real danger with AI is things like millions of AI boyfriends and girlfriends changing the psychology of the next generation, changing the deepest tendencies and structures of the human mind.
He proposes a defence. He calls for a ban on AI personhood. No legal status. No bank accounts. No companies run by machines.
These are terms and conditions. Buyer-side rules, written in advance, that suppliers must accept before they are allowed into the market.
This is the pattern among even the smartest thinkers. Harari is laying out the case for why people should create their own legally enforceable terms and conditions for ethical AI. He walks right up to the start line of Supplierism. Then declares that we need new laws, new regulations and heroic lawmakers to enforce them.
But new laws move at the speed of governments. Suppliers move at the speed of profit. By the time the laws arrive, they are watered down, the product is in a child’s pocket, and the harm is on your balance sheet.
Supplierism does not wait for some new fast and enlightened government to magically appear. It takes on your suppliers using the one thing they need: access to your money. But Supplierism makes that access conditional on good behaviour. With consequences.
Why does this work? Because buyers already have the right to set the terms they need, to govern AI and every other harm their suppliers create. They have always had this right. The only thing missing was the capacity to use it at scale.
Where Ezra Klein misses is more interesting.
Klein (whose work I admire and try not to miss) has spent years arguing that liberalism’s problem is communication. The wrong people are good at podcasts. The right people still talk like the institutions of an era that no longer exists. The algorithm rewards excitement, and excitement rewards cruelty.
All of it is true. None of it fixes the problem.
The solution is not better communicators. It is to change what the algorithm rewards. Engagement is a supplier metric. It was chosen by the supplier, optimized by the supplier, and sold to the supplier’s advertisers. The buyer, meaning you, was never asked what the algorithm should optimize for.
Aggregate what the buyers want instead. Attach it to where the money flows. The algorithm will optimize for something else, not because the supplier had a change of heart, but because the supplier needs your money.
Klein keeps looking for the answer in better elites. The answer is in better buyers.
Harari ends the conversation on finance.
AI will create a new financial system that we will not be able to understand. We will see things happening, like: this company fired me, that company hired me. Why? I have no idea.
Harari is describing a dystopia where AI runs the world and no one can see why. He is right to worry. But the question is no longer whether AI is dangerous. The question is whether it is inevitable. If you believe it is, then the only question left is who the AI works for. A company? A tech billionaire? A government?
Or you.
That is the choice in front of every buyer right now. There will be AI in your bank, in your insurance, in your kid’s classroom, in your doctor’s office. The only question is whose terms it is reading. Who controls the contract.
Harari believes people, working together, should control AI for the good of humans. That likely puts him in your buyer pool. So does the Pope, and the 1.4 billion who follow him. So does nearly everyone who believes in privacy, in human rights, in not handing their child’s mind to whoever bids highest for it.
That is not a fringe. The buyer pool that wants what Harari wants is most of the people and almost all of the money. The group that actually wants the opposite is small: it wants unfettered AI, machines running the economy, no terms at all. And it does not hold the money. Suppliers follow the money. The leverage is on your side.
Harari spent two hours describing the gap. Klein spent two hours not seeing it.
Go to Supplierism.com. State your terms. Then see who you are already standing with.